{
  "id": 271943,
  "title": "Post processing",
  "url": "/competitions/landmark-retrieval-2021/discussion/271943",
  "author_name": "",
  "post_date": "2021-09-13T10:20:51.617045600Z",
  "votes": 1,
  "comment_count": 1,
  "views": 0,
  "content": "<p>How can we get the previous year's test dataset ?? since the test dataset had good number of non landmark images which can be used for post processing and eventually building better model (last year top soln. used the same idea). Plz help me if u have any idea about this</p>",
  "messages": [
    {
      "id": "1511314",
      "postDate": "09/13/2021 10:20:51",
      "content": "<p>How can we get the previous year's test dataset ?? since the test dataset had good number of non landmark images which can be used for post processing and eventually building better model (last year top soln. used the same idea). Plz help me if u have any idea about this</p>",
      "rawMarkdown": "How can we get the previous year's test dataset ?? since the test dataset had good number of non landmark images which can be used for post processing and eventually building better model (last year top soln. used the same idea). Plz help me if u have any idea about this",
      "votes": null
    },
    {
      "id": "1513185",
      "postDate": "09/14/2021 23:38:21",
      "content": "<p>You can find it in the original GLDv2: <a href=\"https://github.com/cvdfoundation/google-landmark\" target=\"_blank\">https://github.com/cvdfoundation/google-landmark</a></p>",
      "rawMarkdown": "You can find it in the original GLDv2: https://github.com/cvdfoundation/google-landmark",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1513185,
      "author_name": "phoenixl",
      "author_url": "",
      "post_date": "09/14/2021 23:38:21",
      "content": "<p>You can find it in the original GLDv2: <a href=\"https://github.com/cvdfoundation/google-landmark\" target=\"_blank\">https://github.com/cvdfoundation/google-landmark</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1511314": "How can we get the previous year's test dataset ?? since the test dataset had good number of non landmark images which can be used for post processing and eventually building better model (last year top soln. used the same idea). Plz help me if u have any idea about this",
    "1513185": "You can find it in the original GLDv2: https://github.com/cvdfoundation/google-landmark"
  },
  "source": "meta"
}